A Study of the Bionomy of the Spanish Pollen Wasp Ceramius hispanicus Dusmet (Hymenoptera, Vespidae, Masarinae): Nesting, Mating, and Flower Associations

17 Pages·2000·8.4 MB·English

by  Volker Mauss

HYM. RES.
J.
Vol. 9(1), 2000, pp. 1-17
A Study of the Bionomy of the Spanish Pollen Wasp Ceramius
hispanicus Dusmet (Hymenoptera, Vespidae, Masarinae): Nesting,
and Flower Associations
Mating,
Volker Mauss and Andreas Muller
(VM) Institut fur Landwirtschaftliche Zoologie und Bienenkunde, Rheinische Friedrich-
Wilhelms-Universitat, Melbweg 42, D-53127 Bonn, Germany, E-mail: [email protected];
(AM) Institut fur Pflanzenwissenschaften, Angewandte Entomologie, ETH, Clausiusstrasse
25/NW, CH-8092 Zurich, Switzerland, E-mail: [email protected]
—
Abstract. Data about the bionomy of the Spanish pollen wasp species Ceramius hispanicusDus-
met are presented for the first time. Following the concept of Gess and Gess (1988) C. hispanicus
can be characterized by the following ethological elements: a. Nest excavated in non-friable soil;
b. Burrow surmounted by a turret from earth extracted from within the burrow; c. Nest possibly
annual; d. Nest with relatively short main shaft, main shaft vertical to sub-vertical, with an ex-
pansion at the bottom of the shaft; e. Main shaft not terminated by a cell; f. Secondary shafts
absent; g. Cells sub-horizontal, in a group to one side, all at different depths; h. A constructed
mud-cell within anexcavated-cell, formed from earth excavated within theburrow; i. Main forage
plants are representatives ofCistaceae, Primulaceae, Lamiaceae, and Fabaceae. Theentranceturret
of the nest is unique in that it is reduced to three distally converging processes. During nest
excavation the females perform pellet dropping flights and use a defined pellet-dropping area.
Water is used to soften the soil. Females alight on the edge of a water source to collect water.
Cell construction rate during this investigation was about 0.5 cells per day. Analysis ofbrood cell
contents revealed Ceramius hispanicus to be polylectic. Cells were found to be provisioned with
pollen originating from six different plant families. The most important pollen sources were He-
lianthemum (Cistaceae), Coris (Primulaceae), several species of Lamiaceae, and Lotus (Fabaceae).
Pollen grains of the same plant taxa were found in crop and rectum of both male and female
wasps. Males patrol and perch at water and to a lesserextent at flowers. Pairings were frequently
observed at water. The daily period of activity lasted about 10 hours. Activity of males at water
was high in the morning and declined during the day, while female activity increased towards
the afternoon.
The genus Ceramius shows a disjunct within the whole genus. For the Palaearc-
distribution with six species-groupsoccur- tic Region comprehensive information is
ring in the Afrotropical Region and two only available for Ceramius tuberculifcr
species-groups in the Palaearctic (Richards Saussure (Ferton 1901, Giraud 1863, 1871,
1962, Gess 1992). For the Afrotropical taxa Mauss 1996a). The knowledge of the bion-
it has been shown that there are distinct omy of the remaining 13 species is very
differences between the species-groups in poor (cf. Fonscolombe 1835, Mauss 1998,
regard to nest construction and flower as- Richards 1963), bionomical data concern-
sociation (Gess 1996, Gess and Gess 1980, ing Ceramius hispanicus Dusmet are com-
1986, 1988, 1990). By contrast, the biono- pletely lacking. Like Ceramius tuberculifcr,
my of the Palaearctic species is insuffi- C. hispanicus belongs to Ceramius species-
ciently known (cf. Gess 1996) to make it group 7 of Richards (cf. Mauss 1996b).
possible at present to draw conclusions However, these species are not very close-
about the evolution of life history traits ly related, i.e. C. hispanicus is not a mem-
Journal of Hymenoptera Research
ber of the C. lusitanicus-complex (Mauss female activity per half hour is the sum of
1996b). Biogeographically C. hispanicus the 20 counts of a complete measuring cy-
seems to be restricted to Central, Eastern cle (category "females at water").
and Southern Spain (cf. Ceballos 1956: 342, Copulations were counted during the
Richards 1962: 107, Mauss unpubl.). complete 1200 s of the measuring cycle ir-
Data concerning habitat, nest architec- respective whether they occurred inside or
A
ture, flower visiting, male activity and outside of the sample area. situation
mating of C. hispanicus are presented here was rated as a "copulatory attempt" when
for the first time and are compared with at least a short struggle on the ground
ethological accounts for other species of could be observed after a male ap-
Ceramius. proached a female and pounced on her.
two thermometers
METHODS Finally, (precision:
0.5°C) were read one of which was situ-
m
Investigations were carried out from 19 ated 0.5 above the ground in the shade
to 26 June 1998. Weather conditions were within a juniper tree (Tair), the other one
good throughout the whole period. Max- was placed on the wet ground with its
imum air temperatures were circa 30 °C. point in the shadow of plants (TBmund).
Sunrise took place at 4h24, sun-transit at Short notes were made about the weather.
llh56 and sunset at 19h28 (Bien in lit.). Radiation conditions were noted applying
Time used is Greenwich Time. Observa- the categories "cloudless" (sun not cov-
tions were made with the aid of close-up ered during the whole period), "hazy"
binoculars (Eschenbach Binoskop) and (sun covered by hazy clouds at least for a
mm
documented using a 35 camera with short period, resulting in half shade) and
mm
a 100 lens (scale up to 1:1) or a 300 "cloudy" (sun covered by clouds at least
mm
lens. for a short period, resulting in shade). The
Activity of males and females at water observed frequencies of the measured be-
was measured on 21 June (from 8h30 to havioural categories were summed up for
18h30) and 24 June (from 7h00 to 18h30). every hour. Then the proportion of activ-
A rectangular sample area of 2m X lm ity for each hour to total activity for the
was marked out with a string. The area day was calculated separately for each cat-
was completely shaded by the ridge of a egory and expressed as a percentage. For
mountain after 17h30. Accuracy of count- statistical analysis, a Chi square test was
ing was improved by use of mechanical performed which was calculated by Aba-
hand-counters. Every half hour a measur- cus Concepts, StatView® Student for Mac-
ing cycle lasting 20 minutes was carried intosh. For 21 June the period from 8h30
out following the sequence: to 12h30 was compared with the period
1. Activity of females measured by from 12h30 to 16h30. For 24 June the pe-
counting the number of females sitting on riod from 6h30 to 12h30 was compared
the ground of the sample area every 30 s with the period from 12h30 to 18h30. For
during a period of 300 s, leading to 10 each category the observed frequencies
counts per period. were tested against the expected equal
2. Activity of males measured by count- distribution.
ing crossings of the string into the sample For nest excavation on 26 June the shaft
area (category "males flying") during a of each nest was completely filled with
period of 600 s and counting landings on Maizena® (fine maize flour) which was in-
a perch within the sample area (category jected with a squeezing bottle. Nests were
"males perching"). carefully excavated afterwards. Measure-
3. Activity of females measured again ments were taken by use of small strips of
uring a period of 300s as described in I.; graph paper, orientation of the cells was
Volume 9, Number 1, 2000
measured with a bearing-compass. All tion towards the southwest. A water
nest cells were collected and stored in a trough which was supplied by a perpetual
refrigerator for three days. Afterwards, all spring non-seasonally was situated at the
cells were measured (external maximal di- upstream end of the narrow part of the
mensions) using a stereo-microscope valley. The water ran away from the
(Wild M3) with a calibrated ocular-micro- trough into a little stream (Fig. 1) that
meter; then they were opened and the dried out after about 100 m. Further
contents were recorded. downstream a few puddles remained at
Flowering plants in the neighbourhood first but they dried out during the obser-
of the nesting site were collected and pre- vation period. Two small ponds were sit-
m
served both dried and in 70% ethanol. uated about 100 upstream of the
They are named according to Tutin et al. trough, the border ofwhich was complete-
(1964-1980). Pollen samples from the nest ly overgrown with rushes (Juncus sp., Jun-
and the alimentary tract of imagines fixed caceae). The whole area was covered by
in Bouin's solution were prepared using sparse montane forest, on calcareous soil,
the method outlined by Westrich and dominated by trees and shrubs of differ-
Schmidt (1986). The different pollen types ent junipers (Juniperus sp., Cupressaceae)
were ascertained under a light microscope (cf. Polunin and Smythies 1973) forming a
at a magnification of 400X or 1000X and Junipereto hemisphaerico-thuriferae sig-
determined to the family or genus level mentum (Rivas-Martinez 1986). The
with the aid of a reference collection con- ground cover was about 70% in the valley
sisting of pollen samples of 500 mainly but decreased markedly uphill where it
Mediterranean plant species including became more rocky and much drier (Fig.
those growing at the nesting site. Exact 1). The following plant species were in
knowledge of the plant species flowering flower: the Lamiaceae Nepeta nepetella L.,
at the study site during nest provisioning Marrubium supinum L., Sideritis spinosa
in some cases allowed pollen determina- Lam. and Thymus zygis L., the Cistaceae
tion down to species level. The percentage Helianthemum apenninum (L.) Mill, and H.
of the different pollen types per brood cell cinereum (Cav.) Pers., the Fabaceae Lotus
was estimated by counting 50 grains at corniculatus L., Coronilla minima L. and
each of 30 loci distributed randomly over species of Ononis, Hippocrepis, Onobrychis,
the cover slip. For each gut sample be- Medicago and Vicia, the Primulaceae Coris
tween 100 and 250 pollen grains were monspeliensis L., the Asteraceae Anacylus
counted. clavatus Pers. and Achillea sp., the Rese-
daceae Reseda lutea L., the Boraginaceae
RESULTS
— Echium vulgare L., the Rosaceae Potentilla
Description of the Imbitnt. A large pop- reptans L. and unidentified species ofBras-
ulation of Ceramius hispanicus was local- sicaceae, Cichorioideae, Convolvulaceae
ized in the Barranco de Zorita (GPS: and Crassulaceae. The area was grazed by
01°26.402 W 40°27.334 N), a small valley sheep and goats. A small cornfield ad-
in the Sierra de Albarracin about 6 km joined in the upper, widened part of the
north of Albarracin in Teruel province, sit- valley. —
uated on the north-east slope of the Val- Nest site. An aggregation of five nests
lejo Largo at an altitude of 1200 m. The was located on a bank of hard, clayey soil
narrow, steeper part of the valley was ori- mixed with some gravel. The bank was
entated from the east-south-east (down- about 20 m long, 2.5 m wide and rose
stream) to the west-north-west (up- above the adjacent terrace by 0.4 m. It was
m
stream); at the upper end it got wider, situated about 50 upstream of the
sloped only gently and changed its direc- ponds and ran from the southwest to the
Journal of Hymenoptera Research
Fig. 1-5. 1, Habitat <>t Ceramius hispanicus at Barranco de Zorita (Prov. Teruel, Spain) covered by sparse
montane forest dominated b\ different junipers. Males patrolled along the stream in thecentre wherefemales
collected water. 2, Male ol ( eramius hispanicus perching on a stone close to the water (glossa visiblebetween
slightly opened mandibles) » Female ol ('.eramius hispanicus Standingon the wet ground at thestream during
water uptake (note extended glossa). 4, Nest No. 1 oi ('animus hispanicus alter excavation on 2h June, 1998
(turret removed/ shaft filled with Maizena" ; s mam shaft; c constructed mud cell). 5, Constructed mud
cells of nest No. 3 of Ccrauiius hispanicus on 29 Iline, 1998 (length ol scale bar 5 mm).
Volume 9, Number 1, 2000
Fig. 6-10. 6, Ceramius hispanicus female during nest excavation. The female had backed out of the entrance
with the soil-pellet held in her mandibles and was about to turn round and start the pellet discard flight.The
nest entrance is surmounted bv three elongated, distallv converging processes (= p; third process mainly
hiddenby the middleone)(seealso [ig. 11)7. femaleofCeramiushispanitus initiating theturretattheentrance
ofthe main shaft (mud pellet supported bv mid-legson the outsidei. B, Female of Ceramius hispanicus placing
a mud pellet on the distal end ofoneofthe processesof the turret. 9, Copulation ofCeramius hispanicus, male
and female grappling on the ground. 10, Copulation of Ceramius hispanicus; male still connected to the female
by its genitalia after it had lost its hold on the thorax of the female that had tried to escape. The situation
lasted for about 180 s.
Journal of Hymenoptera Research
the shaft was obliquely or vertically ori-
entated and was not terminated by a cell.
mm
It widened to 9-10 in diameter in the
region of the brood cells which lay hori-
zontally to sub-horizontally and radiated
out from the main shaft. Secondary shafts
were lacking so that the cell openings
were directly integrated into the wall of
the shaft. The constructed mud cells could
Ftirga.nce11.ofCSecrhaemmiautsichisrpeapnriceussen(tsaeteioanlsoofFigt.he6)n(ewsetakelny- b(Feige.as4i)l.yTsheepacrealtlsedwefrreomeltohnegaatdeh,ermionrgesooirl
dotted area = parts of the nest entrance which are less straight and noticeably wider at mid-
made out of mud by the female; p = elongated, dis- length than before and after it; the inner
tally converging processes probably homologous end was markedly rounded (Fig. 5). Their
with the turret; length ofscale bar 1 mm). outer surface was irregular but more or
less homogenous (Fig. 5); the innersurface
northeast, gently sloping towards the lat- was smooth but dull. Measurements of
m
ter. The nesting area measured about 2 2, each cell and details ofits contents are list-
its ground cover was 40-50%. The mini- ed in Table 1. The cell provision was a
mum distances between the nests varied firm and relatively dry pollen and nectar
from 0.4 to 1.2 m (median 0.7 m). Four loaf which did not adhere—to the wall.
nests were situated on the face exposed to Nest-building behaviour. Initiation of a
the south-east and one was on top of the nestby a female was observed three times.
bank. — The females were flying slowly low over
Nest architecture. All nest entrances the ground. They interrupted their flight
were to some extent hidden under leaves several times to alight on the ground
of plants. The entrance was surmounted which in some instances they scratched
by a low turret which was built of mud with their mandibles. At the future site of
pellets cemented together. It consisted of its nest, each female flew up from the
a short basal ring which continued into ground and performed a circular orienta-
the lining of the shaft towards the base. tion flight, the diameter of which was
Distally, three elongated, converging, col- about lm. Then the female alighted on the
umn-like processes arose obliquely from same spot again and directly started ex-
the basal ring (Figs 6, 11). The processes cavation.
mm
were about 3 long and were situated Excavation was initiated by softening
around that half of the entrance which the soil with a liquid which was appar-
was situated above the outsidebend ofthe ently regurgitated. A pellet of mud was
oblique outset ofthe shaft forming a three- formed by the mandibles; scratching
pronged hood. Basally the processes were movements of the fore- and mid-legs were
separated from each other by gaps of ca. performed in addition. When a pellet had
1 mm. attained about half the size of the head the
The shaft ran obliquely downwardsmmat female flew up with the pellet held be-
an angle of ca. 45° for the first 2-3 tween her mandibles. She flew very rap-
(Fig. 12). Below this it continued down- idly about 0.1 m above the ground to-
wards more or less vertically for 40-60 wards an area situated between 0.3 to 0.7
mm, except in nest No. 2 in which the m away from the nest in the immediate
shaft descended obliquely to avoid stones vicinity of a plant. At the end she hovered
(Fig. 12). This section of the shaft was 5-6 for a short moment, dipped down a few
mm
in diameter. The remaining part of centimetres while dropping the pellet, re-
Volume 9, Number 1, 2000
E
50]
aii
c
T3
E
73
4)
g
re
g
£
ll
en
4C-1
£
sc
u
=
-
u
1
—o
C
£
0j
xO
e
-
c °ou
8 Journal of Hymenoptera Research
Table 1. Details pertaining to the five nests ofCeramius hispanicus excavated on 26June, 1998 at Barranco
de Zorita (measurementsof cells and investigation ofcell content were made on 29 June, 1998).
Volume 9, Number 1, 2000
Table 3. Pollen composition of provision from nine brood cells from four nests of Ceramius hispanicus
collected on 26 June, 1998 at Barranco de Zorita (n = 1500 pollen grains/provision).
10 Journal of Hymenoptera Research
head first. They were able to turn around ternately underneath their body. The for-
inside the nest, but during excavation they etarsi were brought to the mouthparts a
left the entrance a—lways backwards. few times during the process, indicating
Water collection. Females alighted fre- pollen uptake. When on flowers of this
quently at the edge of the little stream or plant species, the females were never ob-
on the damp soil in its immediate vicinity served to insert their heads into the corolla
to collect water. They were never ob- base.
served to settle on the water surface. After The brood cells were provisioned with
landing the females often walked a few a firm and relatively dry loaf composed of
steps forward, stopped, extended theglos- nectar and pollen. The pollen composition
sa and started to take up water, this being of all provision sampled was remarkably
accompanied by vigorous pumping move- similar. All provision contained high per-
ments of the metasoma (Fig. 3). During centages of pollen of Helianthemum apen-
water-collection females often chose spots ninum and H. cinereum (Tab. 3); less im-
on the damp soil or at the edge of the wa- portant pollen sources though well repre-
ter, which were hidden by vegetation. Fe- sented in somebrood cells were Coris mon-
males were observed much less frequently speliensis, four different species of
at the two ponds and the trough than at Lamiaceae, and Lotus corniculatiis. Pollen
the stream. Females visiting the ponds of Convolvulaceae and Crassulaceae oc-
stood on the blades of the rushes during curred in small amounts in some cells,
water uptake; at the trough they held on Likewise, the alimentary tract of the ima-
to the vertical walls a short distance above gines contained pollen grains of Helianthe-
the surface of th—e water. mum, Coris, several species of Lamiaceae,
Forage plants. Flower-visiting records and Lotus; pollen of Asteraceae occurred
for imagines are summarized in Table 2. in addition (Tab. 4)—.
Males visited the flowers of the Lamiaceae Mating behaviour. Males were most fre-
extensively while the observed single vis- quently observed at water. They flew in
its to Helianthemum cinereum and the yel- elliptic flight paths along the stream banks
low Asteraceae were very short. The fe- in a slow, constant flight about 0.1 m
males were observed to visit mainly white above the ground. The most striking fea-
flowering Lamiaceae and Lotus comicula- ture of the flying males was the white col-
tus. Three times females were observed to oration of the clypeus and the mandibles,
change from one plant taxon to another which strongly contrasted with the dark
during a single foraging trip, indicating coloration of the body. In addition, the an-
low flower fidelity. The behavioural pat- tennae, raised at about 45° to the median
tern exhibited on flowers differed remark- axis of the body, showed their conspicu-
ably with the plant taxon. While visiting ously orange-marked curved distal ends,
flowers of Lamiaceae the imagines insert- The patrolling males sometimes interrupt-
ed the mouthparts and the distal parts of ed their flight and alighted on sun-ex-
the head deeply into the corolla. On one posed stones which were situated 0.1 to 1
occasion it was seen with certainty that m (exceptionally 3 m) away from the
the glossa was extended when the head stream. On the perch the males main-
was removed from the flower, indicating tained a characteristic posture. Antennae
nectar uptake. When females alighted on and wings were raised at about 45° to the
flowers of Lotus corniculatiis the alae of the median axis of the body; the head was of-
flower were pressed ventro-laterally. Si- ten slightly raised; and the mandibles
multaneously, the females performed lat- were usually closed, although it was ob-
I movements with the gaster and served a few times that the glossa was
moved the distal parts of the forelegs al- stretched forward slightly (Fig. 2). Perch-
nm